The Panasonic G1HD | GH-1 Video Mode : A Real Video Mode

Ian-T said:
Well...so far it seems like out of the 4 people with the cam you are the hardest on the 24p/1080 footage. But you did say our mileage will vary so maybe Phillip, Hunter or our other friend in Japan has other opinions about it. It's all good.
Click to expand...

I'd love to see what all three feel about the 1080/24P. I'd actually take more comfort in knowing that I'm the only one that thinks it's a bit under-delivered (yes, I know it's a 1500.00 Camera...).

I really hope that someone finds something different and more positive than myself!
 
Ok, there's a debate developing about the GH1's 1080 vs 720. But what I'm interested in knowing right now is what the 1080 compression issue is like compared to other cameras. If you were to shoot a panning shot with say a Canon HV20 (a good yard stick camera for many of us), would a similar shot in 1080 with the GH1 look worse, similar or better?

Kholi -- thanks for your big effort (with Japanese menus!) testing this thing early. I respect your opinion mate, and look forward to reading your conclusions when ready. Many thanks.
 
TrueIndigo said:
Ok, there's a debate developing about the GH1's 1080 vs 720. But what I'm interested in knowing right now is what the 1080 compression issue is like compared to other cameras. If you were to shoot a panning shot with say a Canon HV20 (a good yard stick camera for many of us), would a similar shot in 1080 with the GH1 look worse, similar or better?

Kholi -- thanks for your big effort (with Japanese menus!) testing this thing early. I respect your opinion mate, and look forward to reading your conclusions when ready. Many thanks.
Click to expand...

Compare to an HV20? It's probably the same if not better.

Actually, where's Isaac Brody? He's got an HF10 and he can probably tell you MUCH better than I can. That's a pretty good question, though!

And LOL About the Japanese menus. Hunter sent me some english translations but by the time he did I had already figured it out and commited the menus to memory. At least my brain got a nice workout.

Really soon, seriously.
 
LizaWitz said:
I swear you're coming off as bipolar. One minute its the greatest thing since sliced bread and the next you can't even shoot a brick wall! (No offense meant.)

But I'm just taking comfort that you're comparing it to RED. Lacking as compared to RED is great. I should think that you get better footage for paying 15 times as much!
Click to expand...



The last time I read, Kholi compared the GH1 to being like an EX1 with a large sensor. Correct Kholi?


Which is still by no means short of amazing at this level.
 
Kholi said:
And please remember, just because I think that 1080/24P isn't up to snuff on this camera, not even by HVX200 standards, .
Click to expand...


The 1080P stills from hunter and 1080p vids from Bloom; that level of sensor quality never came out of my HVX200 or any around me. HVX is a solid camera, however the GH1 frequency resolution is evident in the images posted. That said, if you compared it to being an EX1, then by all accounts it is superior to the HVX.

Correct?

These still images from Hunter which are 1080p frame grabs, are way better then the HVX or any HDV camera....

http://www.dvxuser.com/V6/attachment.php?attachmentid=10791&d=1241039195

http://www.dvxuser.com/V6/attachment.php?attachmentid=10789&stc=1&thumb=1&d=1241039195




.
 
Last edited:
Barry_Green said:
Kholi reminded me (probably even in this thread) that I once pointed out that every camcorder release goes through the following stages:

When It's First Announced:
"Oh my Gosh I cannot WAIT this is THE ONE!!!11!11!!!!"

First Footage:
"THIS IS THE ROXXOR!!! I CANNOT WAIT! FINALLY I can now make my MASTERPIECE! Look out, HOllywood!!!"

The General Public gets their hands on it:
"What the hell is wrong with this? Look at this -- this is awful. Look at this intolerable issue ("noise", "split screen", "jello", "skew", "rolling shutter", "vignetting", whatever). Canon/JVC/Panasonic/Red/Sony, HOW could you DO THIS to us? We should start a class-action lawsuit.

Macgregor or Philip Bloom or Disjecta get their hands on it:
HOLEY MOLEY! WOW, now THAT's what I'm talking about! This is the holy grail of all cameras again!

Eventually people learn how to work with it and get great results from it, and those who are in business settle down to turn out fantastic work with whichever tool they've chosen to buy. And those who don't work, but just dream? Oh, they're onto stage 1 of the cycle with the NEXT great announcement (D90, GH1, Scarlet, 5D Mk II, take your pick...)
Click to expand...
Barry,

As so many times over here, you're fully right! Funny observation. :laugh:

Cheers,

E. :-)
 
Kholi said:
What? I SAID 1080/24 is useless to me already. That hasnt' changed. It's a 720/60P camera and it's incredible at 720/60. My opinion surely has NOT changed. I'm sorry if me saying 1080/24P is crap on this camera (...)
Click to expand...
I'm sorry for my lack of humbleness for pointing out again but it does seem that Adam Wilt is a "little bit" :grin: right about interlaced encoding, after all:

http://www.dvxuser.com/V6/showthread.php?p=1594610#post1594610
 
Emanuel, what are you trying to say? Both the PAL and NTSC will have the same type of chroma sampling, 4:2:0 (unlike DV), I think that may be all Adam Wilt is saying.

Of course the PAL version of the GH1 is going to be slightly better in 1080p mode, it is essentially recording progressive, whereas when removing pulldown in NTSC, you are throwing out some redundant info, so less overall bitrate. It would be interesting to see the PAL and NTSC versions compared in this way just to see how much of a difference there is -- might not even be noticeable.
 
codeloss said:
Emanuel, what are you trying to say? Both the PAL and NTSC will have the same type of chroma sampling, 4:2:0 (unlike DV), I think that may be all Adam Wilt is saying.

Of course the PAL version of the GH1 is going to be slightly better in 1080p mode, it is essentially recording progressive, whereas when removing pulldown in NTSC, you are throwing out some redundant info, so less overall bitrate. It would be interesting to see the PAL and NTSC versions compared in this way just to see how much of a difference there is -- might not even be noticeable.
Click to expand...
As matter of fact, the pulldown process won't affect so much in any way other than as both are actually touched in the same way. The point is the interlaced encoding is present on both (1080i 24p or 25p). They're barely the same, with or without pulldown (it doesn't really matter on the topic). The interlaced basis is not the best way to get a progressive outcome. That is one of the weak links of the chain: 1080 i -- Kholi's findings (720/60p) are a natural conclusion.
 
It makes sense to me on a fundamental level, the only one I can understand in all of this compression stuff. LOL.

Perhaps it isn't even a bitrate folly as much as it is an interlaced delivery, like Emanuel states.

Doesn't really matter either way, because it's plenty enough in 720. If that even is the case.
 
codeloss said:
Emanuel, what are you trying to say? Both the PAL and NTSC will have the same type of chroma sampling, 4:2:0 (unlike DV), I think that may be all Adam Wilt is saying.
Click to expand...
Well, no, it's ... okay, I'm saying this without combing through the footage, but -- it's POTENTIALLY a lot worse than that.

Both the NTSC and PAL version are 4:2:0, yes. But there are different TYPES of 4:2:0, and the one that makes the most logical sense is the progressive version. Interlaced 4:2:0 is nasty.

If the PAL version encodes the 25p images as progressive, then they could look noticeably better than the 24p images as carried within a 1080/60i stream. And simple pulldown removal will *NOT* be enough to restore the footage quality -- which may be why folks are complaining about 1080/24p mode?

Here's the deal: 2:3 pulldown is straightforward, one progressive frame gets written to 2 fields, or 3 fields (with that third field being a duplicate of the first). But in interlaced encoding, the 4:2:0 color sampling actually gets spread across fields. So while you may have duplicated detail (luma) in that third field, you might have unique chroma information. So if the 3:2 pulldown just throws away the duplicate fields, it might be throwing out valuable chroma information too.

It's not a problem with 4:1:1 or 4:2:2, since chroma is recorded on every line with those formats. But 4:2:0, and interlaced, = trouble. A 1080/24pN progressive-encoded format would probably produce noticeably superior footage.

If you want to strip out the pulldown out of HDV or AVCHD 24p-over-60i footage, ideally you'd be using an intelligent tool that understands the chroma problem and properly reconstitutes all the chroma information.


It would be interesting to see the PAL and NTSC versions compared in this way just to see how much of a difference there is -- might not even be noticeable.
Click to expand...
Yep, I'd be very interested in seeing that. In theory it sounds like there could be a noticeable difference, but in reality there might be or there might not be.
 
You must log in or register to reply here.
Back
Top